1. Field of the Invention
One or more embodiments of the present invention relate to a system, medium, and method converting colors, and more particularly, to a system, medium, and method converting colors of an input image by estimating the surface spectral reflectance of the image in one or more wavelength ranges by converting the image into a device-independent color signal on the basis of the estimated surface spectral reflectance.
2. Description of the Related Art
Digital image devices for reproducing color images, such as monitors, scanners, and printers, have been developed in order to meet the user's demands for multi-function and high-quality image devices. These digital image devices use different color spaces or color models depending on their purposes. Color models, for example, are divided into device-dependent models and device-independent models. The device-dependent models include, for example, an RGB (red, green, and blue) model, which is an additive dependent model, and a CMYK color model, which is a subtractive color space model. The device-independent models include, for example, a CIE (Commission International de I'Eclairage) LAB model, a CIE XYZ model, and a CIE LUV model.
The CIE LAB model is a method of quantifying a color in spatial coordinates and representing the color by the values of L*(brightness), a*(red-green), and b*(yellow-blue). The CIE XYZ model represents RGB tristimulus values by a set of XYZ, which are different RGB tristimulus values all having positive values. Here, the CMYK color space is typically used in the printer fields and the RGB color space is typically used in the computer monitor fields, such as in the Internet and corresponding output graphics.
Such digital image devices may convert a specific color of an input image and output the converted color, as long as it can output the color of the input image. Therefore, the user may be able see an image that has been converted into a more natural color, for example.
Meanwhile, techniques for converting the color of an input image into a device-independent color include using surface spectral reflectance of a subject. However, devices for measuring the surface spectral reflectance are typically expensive, and may not be suitable for general electric appliances. Therefore, a technique for estimating the surface spectral reflectance on the basis of an input color signal, such as input to a digital camera, has been proposed. However, such a desired technique is not available due to an ill-posed problem in which the order of input data is larger than that of output data, which results in a large estimated errors in the surface spectral reflectance.
Therefore, the inventors herein believe that it would be desirable to have a technique for converting an input color signal into a device-independent color signal with limited error in surface spectral reflectance.